Image reproduction systems, such as high resolution laser recorders, customarily modulate the output of an electro-optic output device, such as a laser, with an image-representative analog input signal, a prescribed characteristic (e.g. amplitude or pulse width) of which varies in accordance with the value of a pixel to be reproduced. In order to faithfully replicate each pixel of an original image, it is necessary that the mechanism for controlling the exposure of a pixel-associated region of the hard copy medium accurately (linearly) track the input data over the dynamic range of the input signal.
Modulation of the optical beam may be accomplished separately of the beam generator itself. For example, the output beam of a stable output device, such as a helium-neon laser, may be modulated by controllably adjusting the amount of light passing though a separate optical attenuator, such as an acousto-optic or electro-optic modulator positioned downstream of the laser. Alternatively, the beam may be modulated by controlling the drive current to the laser itself (for example, by controlling the drive current to an injection laser diode).
Where modulation of the laser output beam is accomplished by a device separate from the laser, the drive circuitry is complex, so that the overall size and cost of the system is necessarily high. Where the (drive current to) the laser is (amplitude) modulated, steps must be taken to compensate for the non-linearity of the operation of the laser over the wide dynamic range of the image signal. An example of such a non-linear output vs. drive current characteristic, for the case of an injection laser diode, is illustrated in FIG. 1 as having a relatively gradual slope similar to that of a light emitting diode at the low end 11 of its operational power range and a relatively steep slope similar to that of an injection laser diode at the upper end 13 of its power range (e.g. from a hundred microwatts to several milliwatts).
To compensate for this slope differential and the knee 15 between the two regions 11 and 13, the laser drive circuit customarily contains a classical high gain feedback loop which monitors the optical output and compares it to the input signal. The difference between these two signals is then amplified and fed to the laser. Unfortunately, while such a feedback loop provides some degree of linearity improvement in the laser output characteristic, its principal control mechanism is the use of a drive amplifier having a very large gain. The requirement of a high gain imposes significant bandwidth requirements on the drive amplifier, which will either be impractical to achieve, or will limit the speed of the optical output waveform. Therefore, since there exists a practical limit between gain and bandwidth (i.e. the gain-bandwidth product of an amplifier has a finite limit), the use of a single feedback loop has a limit on the gain and bandwidth that can be achieved. There are also several practical shortcomings of the component hardware used in the feedback loop.
More particularly, in order to provide compact modularity and prevent the integrity of the monitored output beam from being compromised, the laser emitter device, the feedback photodetector and its associated beam extraction optics are customarily housed in a sealed unit in close physical proximity to each other. Because of this compact arrangement, the temperature of the photodetector is increased by heat given off by the laser emitter, which imparts an unwanted operational variation into the feedback loop. Secondly, the beam extraction optics is such that the spatial cross section of the monitored beam seen by the photodetector changes with laser output intensity. Thus, although employing an amplitude modulated injection laser diode may reduce the size and cost of the system (as contrasted with pulse width modulation of a separate, hardware intensive helium-neon laser unit), compensation for the inherent non-linear operational characteristic of the laser diode has its own set of problems.